Computer Vision for Kinetic Analysis of Lab- and Process-scale Mixing Phenomena

Abstract

A software platform for the computer vision-enabled analysis of mixing phenomena of relevance to process scale-up is described. By bringing new and known time-resolved mixing metrics under one platform, hitherto unavailable comparisons of pixel-derived mixing metrics are exemplified across non-chemical and chemical processes. The analytical methods described are applicable using any camera and across an appreciable range of reactor scales, from development through to process scale-up. A case study in nucleophilic aromatic substitution run on 5L-scale shows how camera and offline concentration measurements can be correlated. In some cases, it can be shown that camera data holds the power to predict reaction progress.